iMedX Thought Leadership Series
FOR HIM LEADERS, CFOs AND AUDIT TEAMS
Author: Jacqui Ellem – HIM Strategic Advisor & Data Manager, iMedX
What 1,668 audited episodes reveal about precision, productivity and revenue integrity.
The value lies not in flagging more episodes, but in flagging the right ones.
Across healthcare, audit teams face the same challenge: rising complexity, growing documentation volumes, workforce shortages and increasing expectations around financial integrity.
The traditional response has been to audit more. More episodes. More flags. More review activity.
Yet our analysis suggests a different question may matter more: Are we auditing the right episodes? A question that plagues all Health Information Services Departments to ensure maximum efficiency and effectiveness of audit program resources.
To answer this question at an industry level, Using Enhance™, iMedX’s audit technology solution, Jacquelyn Ellem our expert HIM Strategic Advisor & Data Manager, analysed audit rule performance across 1,668 inpatient episodes spanning public and private healthcare environments. Rather than measuring only how many opportunities were identified (flags by the rules engine and technology), Jacqui examined which audit rules resulted in validating coding changes, and which consumed auditor effort without producing meaningful outcomes. The working hypothesis was that not all audit rules are created equal.
The findings reveal a critical insight for health services, CFOs and Health Information Managers: the effectiveness of an audit program is determined less by the volume of rules deployed and more by the precision with which those rules identify true coding risk.
Audit precision emerged as an important lens for audit productivity, reimbursement integrity, and sustainable audit performance.
| 1,668 | 41.3% | 63.7% | 58.0% |
| episodes analysed | public hospital validation rate | private hospital validation rate | public rules not validated |
Definition: “Validated” means a rule flag is confirmed and substantiated with auditor review and comments.
In the case of this analysis, 58% of the rules audit teams would validate as being critical for any audit technology, a specific example being is a cellulitis rule targeted at DRG J64B. Whilst auditors considered this rule important because it targets a common clinical cohort and supports review of diagnosis assignment and DRG integrity, analysis showed that 79 episodes were audited, yet only 16 resulted in a DRG change. This equates to approximately 20% of audited episodes leading to a measurable DRG outcome, meaning auditors reviewed a significant volume of records where no DRG adjustment was ultimately identified. Further confirming rule volume alone does not always mean better audit outcomes.
The evolution of coding audit
Many organisations continue to measure audit success through traditional activity metrics:
- Episodes reviewed
- Audit coverage rates
- Number of flags generated
- Number of coding changes identified
While useful, these measures tell only part of the story. An audit program generating hundreds of potential issues may appear productive. However, if a large proportion of those issues are unsupported upon review, significant audit effort is being consumed without delivering corresponding value.
The question shifts from “How many opportunities are we finding?” to “How accurately are we finding meaningful opportunities?”
This distinction becomes increasingly important as health services attempt to do more with finite clinical coding and audit resources.
Looking beyond flag volume
Using Enhance™, iMedX evaluated the performance of coding audit rules across public and private healthcare settings. The objective was not simply to measure coding variation, but to understand whether individual audit rules consistently identified coding opportunities that were supported by auditor review.
Across 1,000 public episodes and 668 private episodes, the same audit concepts often behave very differently. This suggests that audit rule libraries should not be viewed as static assets, but rather in specific refinement based on clinical context, documentation practices and facility-specific casemix patterns.
Public and private hospitals are solving different problems
One of the most important findings was that coding quality and outcomes vary across the different hospital settings. In the public hospital setting for example, audit outcomes trended towards completeness of coded data, episode integrity, and specificity around ventilation to define the complexity of care.
By contrast in the private hospital setting, the focus on the audit outcomes was centered around diagnosis specificity, condition classification, and coding complexity.
| Public environment | Private environment |
| Completeness of coded data | Principal diagnosis selection |
| Episode integrity | Diagnostic specificity |
| Diagnosis and procedure capture | Condition classification refinement |
| Ventilation and clinical complexity checks | Specialty-specific coding complexities |
The same rule may therefore deliver meaningful value in one setting while generating minimal return in another. For executives, this is an important strategic insight: a rule that performs well elsewhere should not automatically be assumed to perform well within hospital operations.
What the measured results show
The public facility applied 88 rules across 281 DRG change assignments. Of these, 116 were supported by auditor comments, representing a 41.3% support rate. Rule coverage reached 62.2% of episodes at a 28.9% DRG change rate. Twenty-nine rules produced no DRG change and a further 22 had no supporting auditor comments, leaving 58.0% of rules with no observed support.
The private facility applied 79 rules across 171 assignments. Of these, 109 were supported, representing a 63.7% support rate. This was more than 20 percentage points higher than the public result, on a lower DRG change rate of 25.6% and with full rule coverage.
The headline is not that one setting outperformed the other. It is that the same rule can behave differently depending on where it runs.
Completeness versus specificity
In the public dataset, value is concentrated in completeness and episode integrity. The PDx/ADx/ACHI and ventilation capture check achieved 84.2% support, open wound with complications achieved 81.0%, and symptom code as principal diagnosis achieved 76.9%. Together, these rules delivered 43 of 116 supported assignments.
In the private dataset, value is created when a rule is concentrated in diagnostic specificity and principal diagnosis selection. Symptom code as principal diagnosis was the largest single opportunity, with 22 assignments and 63.6% support, followed by the adhesions rule family at 61.5% and 54.5%.
Identical rules diverged across both settings. Dietetics intervention returned 14.3% support in public and 57.1% in private. Cellulitis or abscess from an open wound, and albumin or FFP transfusion without diagnosis, each returned 0% in public and 100% in private.
The hidden cost of low-precision auditing
Perhaps the most significant finding was not where audit rules succeeded, but where they did not. The analysis identified groups of rules that repeatedly generated audit activity without corresponding coding outcomes.
In the public dataset, dietetics intervention and the “passed” flag accounted for 53 of 165 unsupported assignments, or 32.1%. Two overlapping hypotension rules, one relating to anesthesia and the other to surgery, each returned 1 in 8 supported assignments. Together they generated 14 unsupported assignments, indicating possible duplicated rule logic or a shared documentation gap rather than two distinct coding risks.
Every unsupported audit flag requires auditor review, clinical interpretation, investigation and administrative effort. When these assignments accumulate, audit teams can spend substantial capacity pursuing issues that ultimately yield no coding change.
From a CFO perspective, this is not simply an operational issue. It is a productivity issue. Resources allocated to reviewing low-value opportunities are unavailable for higher-risk, higher-impact revenue integrity activities.
A new audit metric: supported assignments per auditor hour
Historically, healthcare organisations have focused on auditor coverage, the more episodes being reviewed the better the outcomes have been largely a working theory. But now with more sophisticated technology, the better opportunity is to focus on precision. At iMedX we call this more juice from the same squeeze, to quote our CTO, Mike Smith.
Precision auditing, better directing the auditor hours.
This measure shifts the conversation from activity to value. A smaller number of highly effective audit rules can outperform a larger library containing low-precision logic. The objective becomes optimisation rather than expansion.
More rules are not necessarily better rules. Our findings suggest organisations may be able to improve audit productivity without adding new rules, instead refining, consolidating or retiring underperforming logic.
What high-performing audit programs do differently
The most mature audit program operate as continuous learning systems. Rather than viewing unsupported results as failures, they treat them as intelligence.
- Documentation gaps
- Clinical terminology inconsistencies
- Coding practice variation
- Opportunities for clinical documentation improvement
- Rule design weaknesses
- Use of pattern recognition discoverable from audit outcomes
- Removal of generic rule libraries
These insights can help improve both audit precision and broader clinical documentation performance, creating a feedback loop that strengthens coding quality, documentation quality and reimbursement confidence.
Three actions for HIM leaders, CFOs and audit teams
| 01 | Segment audit strategies by care settingPublic and private organisations face different coding risks and require different rule priorities. Audit strategies should reflect local casemix, funding drivers and documentation patterns rather than rely on generic rule libraries. |
| 02 | Measure precision as well as coverageIntroduce metrics that evaluate audit yield, supported outcomes and auditor productivity alongside traditional coverage measures. |
| 03 | Treat unsupported results as intelligencePersistently unsupported rules can reveal broader system issues. Feed these findings into clinical documentation improvement, education and future rule optimisation. |
Key Takeaway
Healthcare organisations are under increasing pressure to improve revenue integrity, strengthen coding quality and maximise the value of limited workforce resources.
The future of auditing will not be defined by organisations that generate the most audit flags. It will be defined by those that identify the right opportunities with the highest level of precision.
Our analysis demonstrates that audit performance is highly contextual. The same rule can perform very differently depending on the environment in which it is deployed, the documentation culture of the organisation and the nature of the patient population being served.
When evaluating any technology solution, it’s not necessarily the volume of rules available as the assurance net, it’s the application of these rules and their specificity across the various settings. Results are also highly contextual on the experience level, and precision of human in the loop auditing teams to know where to find the additional complexity to capture additional funding value.
Move beyond volume. Measure precision. Continuously learn.
The strongest audit programs are not those that raise the most flags. They are the ones that continuously improve from every result, whether supported or unsupported.
About Enhance™
Enhance™ is iMedX’s audit module used in this analysis to evaluate coding audit rule performance across public and private healthcare settings. It is designed to give auditing teams pinpoint accuracy in their audit programs, and when integrated and adopted at the point of coding validates clinical coding effectiveness for funding assurance.